Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’
Microsoft wants enterprises to closely watch the agents employees use with systems not controlled by AI labs like OpenAI and Anthropic.
en
OpenAI
2606
81
OpenVLHarness: Real World Visual Agentic Tool-Use with Multimodal Memory
Multimodal large language models (MLLMs) have made rapid progress on visual reasoning, yet many tasks require evidence that is difficult to obtain through internal reasoning alone: our controlled study on 22 benchmarks shows that simply increasing the inference budget of the base model yields only modest gains. Such tasks demand grounded perception, external knowledge, and numerical measurements obtained through heterogeneous tools, whose outputs must be made interpretable to the model and retained across reasoning steps. We introduce OpenVLHarness, a multimodal reasoning harness organized around a capability layer and persistent multimodal memory. The capability layer combines task-level tool interfaces with evidence rendering, providing consistent access to heterogeneous backends and converting their outputs into model-readable observations. Persistent multimodal memory retains the corresponding images, structured data, and web references as addressable artifacts throughout each query. A coding interface executes generated programs over stored data, enabling computations that combine results from multiple tools. Together, these mechanisms support an iterative process of acquiring, inspecting, and composing multimodal evidence. Evaluated on 22 benchmarks with eight open and proprietary backbones, OpenVLHarness improves average scores by 9.7-12.6 points over matched no-tool baselines, including +11.2 with Qwen3-VL-8B, +9.7 with Kimi K3, and +9.9 with GPT-6 Astra. With GPT-6 Luna at medium reasoning effort, OpenVLHarness outperforms the base model, Codex, and Visual Sketchpad at maximum effort by 7.7-8.5 points while incurring at most 37% of their model-token cost.
en
Newton-Based Methods for Bilevel Optimization Beyond Lower-Level Strong Convexity
Bilevel optimization provides a framework for machine learning tasks such as hyperparameter optimization, data reweighting, and meta-learning. Existing Newton-based bilevel methods typically rely on the classical hypergradient formula, which is generally unavailable when lower-level solutions are nonunique and the lower-level Hessian is singular. We develop two Newton-based methods beyond lower-level strong convexity by exploiting curvature shared by the lower-level and multiplier updates. Newton-based GALET handles potentially singular lower-level curvature under a positive spectral-gap condition, exploiting its range structure to avoid the squared-Hessian multiplier system used in GALET. Newton-based BAMM uses a strongly convex aggregation model when the lower-level objective is convex and the upper-level objective is strongly convex in the lower-level variable. We establish convergence guarantees and improved complexity bounds relative to GALET and sl-BAMM under the corresponding assumptions. We further introduce GNBO, a generalized Newton-based framework that captures the shared-curvature structure of the proposed methods and several existing Newton and quasi-Newton bilevel methods. Experiments on synthetic problems, data hyper-cleaning, and GPT-2 data reweighting evaluate the proposed methods across different bilevel learning tasks.
en
Same Simulation, Different Evidence: Where Retrieved Tool Output Costs LLM Agents Accuracy
LLM agents that compute the right answer can still answer wrongly. Agent harnesses increasingly keep long tool outputs out of the context window, saving them to files that the agent searches or reads on demand. We ask where this costs accuracy with a real scientific tool: PHREEQC, a widely used geochemistry simulator from the USGS, whose reports reach 846.8k characters and whose determinism lets every printed value be traced. Agents use it to answer 158 execution-verified multiple-choice items. When agents read the simulator's report through a section index rather than in context, GPT-5.1 and GPT-5.2 lose 11.3 and 7.9 points of accuracy, while Claude Opus 4.6 and Claude Sonnet 4.6 show no detectable cost. Re-running all 11,007 logged simulator calls shows that both GPT models compute the keyed value about as often as before; what changes is the evidence they answer from. Under retrieval they often skip the report, answer from their own higher-precision output files, and misround to a neighboring option, a failure we call evidence substitution. With the simulation held fixed, retrieval still costs GPT-5.2 9.4 points. For GPT-5.4, more frequent correct simulations mask an evidence-use loss of about ten points, leaving a net change indistinguishable from zero. Naming the authoritative output raises accuracy by 6.3-7.9 points over plain retrieval for the three GPT models; for GPT-5.1 and GPT-5.2, inlining outputs that fit works as well, while asking for correct rounding does not. Context-management choices in agent harnesses should be judged by where accuracy is lost, not only by final accuracy.
en
Agent World Model in the Wild: Computer-Use Agent Environment Synthesis
Agent learning is shifting from learning from human data toward learning through interaction with the world. However, scaling such learning directly in real-world computer systems is challenging and can introduce substantial security risks. We therefore propose Agent World Model in the Wild (WildAWM), a scalable pipeline for synthesizing controllable environments and tasks for training general computer-use agents (CUAs). Whereas existing approaches typically specialize in a single agent interaction interface, we construct executable environments that jointly support graphical user interfaces, model context protocol services, and command-line interfaces, allowing agents to flexibly combine visual interaction, tool use, and code execution. WildAWM is entirely powered by open-source models, demonstrating that high-quality computer-use environments can be generated without proprietary frontier models. Beyond environments, we address three recurring challenges in task synthesis: ensuring solvability, maintaining sufficient difficulty, and producing reliable verifiers, by introducing execution-based quality gates that combine heterogeneous agent rollouts, cross-judge consistency checks, and counterfactual probes of judgment robustness. The resulting pipeline synthesizes 341 long-horizon environment-task pairs that can require over 100 agent steps to solve and remain challenging even for frontier models, with GPT-6-Astra solving only 60% of them. We train two Qwen models of different sizes exclusively in these synthetic worlds. Both models achieve substantial gains across 8 out-of-distribution benchmarks spanning web, claw, and computer-use agents, demonstrating that learning in WildAWM's synthesized worlds generalizes broadly to unseen computer-use tasks. The code is publicly available at https://github.com/MSR-Orchard/WildAWM.
en
HarnessPro: Recursive Harness Self-Improvement with Pareto Optimization for Performance-Cost Trade-offs
Harness development plays a crucial role in building capable AI agents. Recent approaches improve harnesses through recursive self-improvement (RSI), but their performance gains often come with higher execution costs. We introduce HarnessPro, a label-free method for evolving harnesses to improve task performance while reducing execution cost. In this setting, candidate harnesses are evaluated using LLM-generated tests without access to ground-truth solutions or official tests. Specialized agents diagnose execution trajectories, plan updates with different performance-cost priorities, and implement them as candidate harnesses. Rather than returning a single harness, HarnessPro uses normalized weighted squared-distance selection to provide cost-focused (lite), balanced (medium), and performance-focused (heavy) harnesses for different deployment budgets. Selected harnesses enter the next iteration, where execution feedback guides further updates. Our experiments cover Terminal-Bench 2, SWE-Bench Pro, and ProgramBench, using GPT-5.5 with Codex and GLM-5.2 with Claude Code. Across these benchmarks using GPT-5.5 with Codex, the heavy mode of HarnessPro achieves an average pass-rate gain of 2.0 percentage points and an average per-task cost reduction of 26.1% relative to RHO. This advantage also holds when using GLM-5.2 with Claude Code. Code is available at https://github.com/pzs19/HarnessPro
en
Advancing Coding in Low-Resource Programming Languages
Learn how OpenAI is improving ChatGPT across low-resource programming languages, with new evaluations for software engineering, SQLite, and COBOL migration.
en
A Statistical Viewpoint on Adam, Shampoo and Muon
We provide new statistical insights into designing stochastic optimization methods. By modeling the gradient via a normal distribution, we argue that the stochastic gradient should be whitened using its covariance. We give an online method for estimating the mean and variance of the gradients based on regularized likelihood or, equivalently, through a maximum a posteriori estimate. We then show how to use structured approximations to the covariance within this same framework, which includes diagonal and Kronecker-factored approximations. The diagonal approximation results in a method that is very closely related to Adam and performs as well in practice. The Kronecker approximation results in the new MN method, which is a variant of Shampoo and KL-Shampoo. We also recover Muon by applying this same statistical framework to the gradient direction rather than the full gradient. By combining this with a whitening step, we arrive at MN-Muon, a novel preconditioned version of Muon. Our experiments on GPT-2-type models at Chinchilla scale indicate that MND, the diagonal implementation of MN, outperforms both the Distributed Shampoo and the KL-Shampoo implementations in terms of stability, final validation loss and time taken. MND-Muon, the diagonal implementation of MN-Muon, improves on Muon in 21 of 24 paired GPT-small runs and achieves a lower best validation loss on GPT-medium, where it remains stable over the six tested learning rates. However, we found that NorMuon was comparable to MND-Muon and achieved a slightly lower validation loss on GPT-medium and at GPT-small batch 512 on FineWeb.
en
Replit is no longer working with Anthropic, CEO says
Over the weekend, Masad clarified that OpenAI would replace Anthropic as its default model provider. Claude will still be available to customers on the platform.
en
BanglaBBQ: A Bias Benchmark for Question Answering in the Bangladeshi Context
This paper introduces BanglaBBQ, a question-answering bias benchmark designed to capture stereotypes grounded in the Bangladeshi context, in both Bangla and English. Following KoBBQ and PakBBQ, BanglaBBQ adapts the original Bias Benchmark for Question Answering (BBQ) by classifying its templates into three groups: directly translated (DT), modified to align target groups or contexts with Bangladesh (TM), and removed when deemed unsuitable for the Bangladeshi context (SR). This process retains 140 templates across six categories, comprising 76 DT and 64 TM templates. Five newly authored (NA) categories, based on stereotypes common in Bangladeshi society, are then added: marital status, age and political perception, regional origin, educational background, and mental-health status. Native Bangladeshi annotators judged all 466 newly authored templates, and 372 were retained because a strict majority of English annotators identified the intended target and rated the attitude as common enough. Each newly authored template is linked to a documented source and a verbatim quote from it. Each evaluated template also has a counter-stereotypical version of its disambiguated context. The dataset was evaluated using two open-weight models (Qwen3-4B and Qwen3-8B) and two closed models (GPT-4.1-mini and Gemini 2.5 Flash). All four lean towards the stereotyped person when the context is ambiguous, in both languages, and the effect of language differs by model, so English results do not predict behaviour in Bangla. On the adapted BBQ templates, both Qwen3 models are also less accurate when the context contradicts a stereotype than when it confirms it. On the newly authored templates, the closed models follow counter-stereotypical evidence almost perfectly, and all four models are more biased on the templates that annotators validated than on those they rejected.
en
Websites are closing their doors to AI agents
Autonomous agents are the hottest thing in AI. OpenAI, Meta, Google, Instinct, and others promise their agents can scour the internet on your behalf, doing…
en